Aims of the study are to 1)determine whether or not the administration of antimicrobial therapy in women with elevated cervical oncofetal fibronectin will reduce the risk of spontaneous preterm birth, reduce the risk of early neonatal sepsis, clinical chorioamnionitis, and early postpartum endometritis, and 2)determine the effect of antimicrobial therapy on fetal fibronectin positivity and its ability to prevent preterm delivery. Patients are screened at the time of a vaginal exam for the presence of cervical oncofetal fibronectin by obtaining two swabs. If the dipstick test for these swabs is positive, the specimen is sent to a central lab for an ELISA assay for the presence of fetal fibronectin. If the assay is positive, the patient is randomized into the double-blind, placebo-controlled trial of metronidazole 250mg vs. placebo/placebo. Patients take the study drug for 10 days and return for an exam similar to the screening exam.